Journal of the American College of Cardiology
© 2000 by the American College of Cardiology
Published by Elsevier Science Inc.
Vol. 35, No. 2, 2000
ISSN 0735-1097/00/$20.00
PII S0735-1097(99)00541-0
Garlic Powder, Effect on Plasma
Lipids, Postprandial Lipemia,
Low-Density Lipoprotein Particle
Size, High-Density Lipoprotein
Subclass Distribution and Lipoprotein(a)
H. Robert Superko, MD, FACC,* Ronald M. Krauss, MD†
San Mateo and Berkeley, California
OBJECTIVES
To test the hypothesis that a garlic supplement alters plasma lipoproteins, postprandial
lipemia, low-density lipoprotein (LDL) size and high-density lipoprotein (HDL) subclass
distribution differently in 50 moderately hypercholesterolemic subjects classified as LDL
subclass pattern A or B.
BACKGROUND Garlic has been variably reported to reduce or not affect plasma cholesterol values.
Low-density lipoprotein pattern B is a common inherited disorder of lipoprotein metabolism
that has been shown to have a significantly greater response to several lipid lowering
treatments including low fat diet when compared with LDL pattern A individuals.
METHODS
A double blind, randomized, placebo controlled trial in an outpatient lipid research clinic was
performed and included fifty moderately hypercholesterolemic subjects (mean LDL cholesterol ⫽
166 ⫾ 22 mg/dl) classified as LDL subclass pattern A (predominantly large LDL, n ⫽ 22) or B
(predominantly small LDL, n ⫽ 28). Following a two-month stabilization period, subjects were
randomly assigned to a placebo or 300 mg three times a day of a standardized garlic tablet for three
months.
RESULTS
For all subjects, LDL pattern A and B subjects combined, garlic treatment for three months
resulted in no significant change in total cholesterol, LDL cholesterol, HDL cholesterol,
HDL subclass distribution, postprandial triglycerides, apolipoprotein B, lipoprotein (a)
(Lp[a]), LDL peak particle diameter or LDL subclass distribution. There was no significant
difference in response for the same parameters among subjects classified as LDL pattern A or
B with the exception of significantly greater (p ⫽ 0.01) reduction in mean peak particle
diameter in pattern A subjects treated with either garlic or placebo. There was no significant
change in LDL subclass distribution.
CONCLUSIONS This investigation confirms that garlic therapy has no effect on major plasma lipoproteins and
further, that it has no impact on HDL subclasses, Lp(a), apolipoprotein B, postprandial
triglycerides or LDL subclass distribution. Garlic may have a greater effect on LDL particle
diameter in LDL pattern A compared with pattern B subjects. This difference was not
reflected in other plasma lipid measurements. (J Am Coll Cardiol 2000;35:321– 6) © 2000
by the American College of Cardiology
Garlic (allium sativum) has been reported to have a beneficial effect on several cardiovascular risk parameters includFrom the *Cholesterol, Genetics and Heart Disease Institute, Berkeley Heart Lab,
San Mateo, California; and the †Department of Molecular and Nuclear Medicine,
Life Sciences Division, Lawrence Berkeley National Laboratory, University of
California, Berkeley, California. This study was supported by the Cholesterol,
Genetics and Heart Disease Institute and a grant from Kwai Pharmaceuticals and
NIH Program Project Grant HL 18574 from the National Heart, Lung and Blood
Institute of the National Institutes of Health and was conducted at the Cholesterol,
Genetics and Heart Disease Institute and the Lawrence Berkeley National Laboratory
(Department of Energy contract DE-AC03-76SF00098 to the University of California), Berkeley, California.
Manuscript received May 4, 1999; revised manuscript received August 16, 1999,
accepted October 18, 1999.
ing plasma lipoproteins and postprandial lipemia (1– 4).
However, a recent double-blind, randomized, placebo controlled trial has reported no effect of a garlic oil preparation
on serum lipoprotein values (5). Low-density lipoprotein
(LDL) subclass pattern B, compared with pattern A, is an
atherogenic lipoprotein profile that increases cardiovascular
risk three-fold and is associated with an abundance of small
LDL, moderately reduced high-density lipoprotein cholesterol (HDLC), moderately elevated triglycerides and increased postprandial lipemia (6,7). Cholesterol lowering
therapies, including diet, niacin, bile acid binding resins,
gemfibrozil and hormone replacement therapy, but not
HMGCoA reductase inhibitor therapy, have been reported
322
Superko et al.
Garlic, Lipids, LDL and HDL Subclass Distribution
Abbreviations
Apo
⫽
BMI ⫽
C
⫽
HDL ⫽
HDLC ⫽
LDL ⫽
LDLC ⫽
Lp(a) ⫽
t.i.d.
⫽
VLDL ⫽
and Acronyms
apolipoprotein
body mass index
cholesterol
high-density lipoprotein
high-density lipoprotein cholesterol
low-density lipoprotein
low-density lipoprotein cholesterol
lipoprotein (a)
times a day
very-low-density lipoprotein
to have a significantly different effect in patients classified as
LDL pattern A versus B (8 –11). Part of the confusion
surrounding the effect of garlic on lipoprotein values may be
due to a differential lipoprotein response in LDL pattern A
versus B subjects so that the poorly responsive group blunts
the effect in the entire group and masks a significant change
in the responsive group. This issue is of clinical importance
because, in four coronary arteriographic regression trials,
LDL subclass distribution has been reported to be associated with arteriographic change following treatment (12–
15).
In order to determine if garlic had a differential effect on
lipoprotein response, similar to the effect of low fat diet,
niacin, bile acid binding resins, gemfibrozil and hormone
replacement therapy on LDL particle size, LDL subclass
distribution and HDL subclass distribution, we investigated
the effect of standardized garlic tablets on 50 subjects in a
double blind, randomized, placebo controlled trial, with
determination of LDL and HDL subclass pattern and
distribution. Other measures of lipoprotein metabolism,
including lipoprotein(a) (Lp[a]), apolipoprotein B (Apo B)
and postprandial triglyceride response were measured in
order to determine if garlic therapy impacts these parameters which may not be apparent from measures of routine
plasma lipids.
METHODS
Subjects. Fifty subjects (mean age 53 ⫾ 10 years, weight
163 ⫾ 30 lbs) were recruited on the basis of low-density
lipoprotein cholesterol (LDLC) ⬎150 mg/dl and ⬍200
mg/dl and triglycerides ⬍300 mg/dl. Informed consent was
obtained from all participants. Subjects were excluded if
they were shown to have heterozygous familial hypercholesterolemia, a systemic illness that could affect blood lipids,
body weight greater than 30% ideal, use of lipid lowering
drugs in the preceding two months or other medications
known to alter blood lipids. Following this phase, subjects
were randomized in a double-blind manner, to placebo or
garlic, 300 mg three times a day (t.i.d.) Kwai garlic tablets
(Lichtwer Pharma, Pittsburgh, Pennsylvania) for 12 weeks.
This dose has previously been reported to result in a
significant 8% reduction in LDLC (16).
JACC Vol. 35, No. 2, 2000
February 2000:321–6
Diet analysis. All subjects were stabilized on an American
Heart Association Step I diet for at least two weeks before
entry into the protocol. Three-day diet records were obtained at baseline before randomization and at 12 weeks in
the double blind phase. Food records were analyzed using
Nutritionist III software (17).
Postprandial load. Each subject was fed a laboratory prepared 50% fat meal matched to body surface area as
previously described (18).
Laboratory. Blood samples were obtained following a 16 h
fast and advice to avoid alcohol for the 48 h before the blood
draw. Triglyceride, total cholesterol, LDLC and HDLC
were determined by enzymatic methods and a modified
heparin-2M MnCl2 procedure to precipitate very-lowdensity lipoproteins (VLDL) and LDL (19). These measurements were monitored in the Centers for Disease
Control-National Heart Lung and Blood Institute Lipoprotein Standardization Program (20). The apo B assay
was carried out by a competitive enzyme-linked immunoassay procedure using well-characterized and specific monoclonal antibodies (21). Lipoprotein(a) concentration was
measured with an enzyme-linked immunosorbent assay kit
[Macra Lp(a) Terumo Diagnostics Division] as described
previously (22). This assay uses a monoclonal capture
antibody immunospecific to Apo (a) and a peroxidaseconjugated polyclonal detection antibody with recognition
of the entire Lp(a) molecule. Internal quality assurance for
apolipoproteins was monitored at two levels for each analyte
on an ongoing basis using specifically prepared frozen pools.
Apolipoprotein E isoforms were determined by isoelectric
focusing of VLDL aoplipoproteins and phenotypes designated according to recommended nomenclature (23,24).
Throughout the period in which all apolipoprotein measurements were performed, the laboratory participated in
the CDC-IUIS Apolipoprotein standardization program
(25).
Identification and densitometric measurements of LDL
species were carried out using Pharmacia PAA 2/16%
gradient gels as described previously (26,27). Criteria described previously (28) were used to classify the LDL
subclass pattern as either pattern A, which had the predominant peak ⬎262 angstrom with skewing to the right, or
pattern B, which had the predominant peak ⬍255 angstrom
with skewing to the left. In the size range between 255 and
262 angstroms (15% to 20% of subjects), peaks can be
symmetric, broad or multimodal and result in an intermediate LDL subclass pattern. For purposes of analysis,
subjects classified as the intermediate pattern were included
in the LDL pattern B group. Percent LDL distribution in
seven regions (I, IIa, IIb, IIIa, IIIb, IVa, IVb) was determined. Region IIIa ⫹ IIIb correlates with the atherogenic
region (Sf3–5) on analytic ultracentrifugation.
High-density lipoprotein subclass distribution was determined by gradient gel electrophoresis of HDL and was
performed as previously described (28). Electrophoretic
Superko et al.
Garlic, Lipids, LDL and HDL Subclass Distribution
JACC Vol. 35, No. 2, 2000
February 2000:321–6
323
Table 1. Baseline Mean Group Variables
Triglyceride
TG pp
TC
LDLC
HDLC
LDL dia (A)
LDL IIIa ⫹ IIIb%
HDL2b%
HDL2a%
HDL3a%
HDL3b%
HDL3c%
Apo B
Lp(a)
All
Subjects
(n ⴝ 50)
Garlic
(n ⴝ 25)
Placebo
(n ⴝ 25)
p
137 ⫾ 59
304 ⫾ 146
245 ⫾ 26
166 ⫾ 22
51.6 ⫾ 12.1
256.1 ⫾ 9.2
19.6 ⫾ 10.3
16.5 ⫾ 7.3
23.0 ⫾ 4.7
35.4 ⫾ 6.0
18.2 ⫾ 5.8
6.7 ⫾ 5.6
126 ⫾ 27
27.6 ⫾ 28.5
145 ⫾ 54
320 ⫾ 135
250 ⫾ 29
169 ⫾ 25
51.3 ⫾ 11.5
255.8 ⫾ 9.2
20.2 ⫾ 10.7
14.8 ⫾ 7.1
22.9 ⫾ 3.9
37.7 ⫾ 5.5
19.3 ⫾ 6.1
6.1 ⫾ 3.6
125 ⫾ 23
28.8 ⫾ 33.5
128 ⫾ 63
287 ⫾ 156
239 ⫾ 23
162 ⫾ 18
51.9 ⫾ 12.8
256.4 ⫾ 9.3
19.0 ⫾ 10.1
18.2 ⫾ 7.2
24.0 ⫾ 5.3
33.2 ⫾ 5.7
17.1 ⫾ 5.3
7.4 ⫾ 5.0
127 ⫾ 32
26.3 ⫾ 22.9
0.31
0.43
0.17
0.24
0.86
0.80
0.69
0.10
0.14
0.006
0.18
0.29
0.75
0.76
Mean (⫾SD) baseline triglyceride (mg/dl), lipoprotein cholesterol (mg/dl), LDL peak particle diameter (Angstom), apolipoprotein B (mg/dl) and percent distribution in HDL2b, HDL2a, HDL3a, HDL3b and HDL3c values. LDLC ⫽ low density
lipoprotein cholesterol; HDLC ⫽ high density lipoprotein cholesterol; Apo ⫽ apolipoprotein.
TG pp ⫽ triglycerides 4 h after standard oral fat load.
bands representing the HDL subspecies HDL-2b, HDL2a, HDL-3a, HDL-3b and HDL-3c are identified and
densitometrically scanned using a computer assisted scanning procedure developed at the Donner Laboratory, University of California, Berkeley. The HDL region is by
definition, a density (d) less than 1.21 g/ml, which is then
stained following GGE.
Statistics. Statistical analysis involved a Student t test to
test for significance of difference between groups and change
from baseline values within groups. Analysis of variance was
used to test for the significance of difference in response
between groups. Statistical tests were performed using
Statview software (29).
RESULTS
At randomization, there were no significant differences
between treatment groups for lipid measurements including
triglycerides, postprandial triglyceride response, total cholesterol, LDL cholesterol, LDL peak particle diameter,
LDL subclass distribution, HDLC, apo B, and Lp(a)
(Table 1). There was no significant difference in body mass
index (BMI), blood pressure or diet variables including total
calories, percent calories from total fat, saturated fat, carbohydrate, alcohol, grams of cholesterol or grams of soluble
and insoluble fiber. At baseline, there was no significant
difference between the group randomized to garlic or
placebo for any of the lipoprotein measurements, BMI or
diet analysis with the exception of a significantly higher
HDL3a% distribution (p ⬍ 0.01) in the garlic compared
with placebo group (Table 1). At baseline, previously
described differences between LDL pattern A (n ⫽ 28) and
B (n ⫽ 22) subjects, were observed. Specifically, fasting
triglycerides (112 ⫾ 50 mg/dl pattern A, 169 ⫾ 54 mg/dl
pattern B) and postprandial triglyceride change (96 ⫾ 59
mg/dl pattern A, 173 ⫾ 85 mg/dl pattern B) were significantly higher (p ⬍ 0.001), HDLC (55.8 ⫾ 12.7 mg/dl
pattern A, 46.3 ⫾ 8.9 mg/dl pattern B) significantly lower
(p ⬍ 0.005), LDL peak particle diameter (268 ⫾ 6A pattern
A, 255 ⫾ 6A pattern B) significantly smaller (p ⬍ 0.0001)
and percent distribution in LDL IIIa ⫹ IIIb (12.2 ⫹ 3.5%
pattern A, 28.8 ⫹ 8.4% pattern B) significantly higher (p ⬍
0.0001) in the LDL subclass pattern B subjects compared
with pattern A (3). In pattern B subjects, percent distribution in HDL2b was significantly lower (p ⬍ 0.01) and
distribution in HDL3b significantly higher (p ⬍ 0.01) than
pattern A subjects. Body mass index (24.5 ⫾ 3.9 pattern A,
26.6 ⫾ 4.3 pattern B) was higher (p ⫽ 0.07) in pattern B
subjects.
In the 50 participants, 12 had an Apo E 4/3 phenotype (5
garlic, 7 placebo), one an Apo E 3/2 phenotype (placebo)
and one an Apo E 4/2 phenotype (placebo). The remaining
36 participants had Apo E 3/3 phenotype.
There was no significant effect of therapy on change in
levels of fasting triglycerides, postprandial triglycerides, total
cholesterol, LDL cholesterol, LDL peak particle diameter,
LDL subclass distribution, HDL cholesterol, HDL subclass
distribution, Apo B or Lp(a) (Table 2). There was no
significant between-group differences for change in BMI,
systolic or diastolic blood pressure or diet variables between
the garlic versus placebo group.
Within the pattern A and pattern B groups, comparison
of garlic versus placebo treatment differences revealed no
significant differences for change in fasting triglycerides,
postprandial triglycerides, LDL cholesterol, LDL subclass
distribution, HDL cholesterol, Lp(a), Apo B, postprandial
324
Superko et al.
Garlic, Lipids, LDL and HDL Subclass Distribution
JACC Vol. 35, No. 2, 2000
February 2000:321–6
Table 2. Mean Change Values (mg/dl, ⫾SD) for All Subjects Randomized to the Placebo or
Garlic Group
Triglycerides
TG pp
Total cholesterol
LDL cholesterol
HDL cholesterol
LDL diameter (A)
LDL IIIa ⫹ b%
Apo B
Lp(a)
HDL2a%
HDL2b%
HDL3a%
HDL3b%
HDL3c%
Garlic
(n ⴝ 25)
Placebo
(n ⴝ 25)
p
⫺4.3 ⫾ 43.8
⫺49.6 ⫾ 73.6
⫺2.5 ⫾ 26.4
⫺1.7 ⫾ 25.5
⫺0.1 ⫾ 6.6
⫺1.7 ⫾ 7.8
3.9 ⫾ 15.0
⫺0.7 ⫾ 25.3
⫺2.5 ⫾ 15.4
0.4 ⫾ 4.5
1.0 ⫾ 4.1
⫺1.8 ⫾ 6.1
⫺0.2 ⫾ 4.5
0.6 ⫾ 6.2
21.5 ⫾ 96.8
⫺24.8 ⫾ 87.1
1.2 ⫾ 19.0
⫺3.2 ⫾ 14.9
0.2 ⫾ 8.9
⫺3.7 ⫾ 7.0
4.3 ⫾ 10.8
⫺7.0 ⫾ 31.8
⫺0.7 ⫾ 8.2
0.8 ⫾ 5.6
⫺0.6 ⫾ 7.9
1.2 ⫾ 7.7
⫺0.6 ⫾ 7.1
⫺0.7 ⫾ 8.0
0.23
0.28
0.57
0.80
0.91
0.35
0.91
0.45
0.61
0.76
0.36
0.13
0.78
0.52
Low density lipoprotein diameter (A) in angstroms. p is the significance of the difference between the change in the placebo
group compared with the garlic group.
Apo ⫽ apolipoprotein; HDL ⫽ high-density lipoprotein; LDL ⫽ low-density lipoprotein; Lp(a) ⫽ lipoprotein(a); TG ⫽
triglyceride.
triglycerides, HDL subclass, as well as diet records, systolic
or diastolic blood pressure or BMI. The only significant
difference (p ⫽ 0.01) was a greater reduction in LDL peak
particle diameter in pattern A subjects in both the garlic and
placebo groups, compared with pattern B subjects (Table 3).
There was no difference in the percent distribution in the
seven LDL regions. There was no significant difference
between the pattern A subjects treated with placebo versus
garlic (p ⫽ 0.92). Comparison of change in the LDL
pattern A and B groups treated with garlic revealed no
significant differences with the exception of significantly
greater (p ⫽ 0.02) LDL peak particle diameter in pattern B
compared with A subjects. Regression to the mean is
unlikely since within the placebo group there was no
significant differences in diet, weight or lipoprotein variables
between LDL pattern A and B subjects.
DISCUSSION
New findings. This double blind, placebo controlled trial
in 50 subjects found that daily garlic supplementation had
no significant effect on plasma lipids, LDL peak particle
diameter, LDL subclass distribution, HDL subclass distribution, Apo B, Lp(a), postprandial lipemia and systolic and
Table 3. Mean Change (mg/dl) by LDL Pattern After Placebo or Garlic Treatment
Pattern A
Triglycerides
TG pp
Total cholesterol
LDL cholesterol
HDL cholesterol
LDL diameter (A)
LDL IIIa ⫹ b%
Apo B
Lp(a)
HDL2a%
HDL2b%
HDL3a%
HDL3b%
HDL3c%
Pattern B
Garlic
(n ⴝ 12)
Placebo
(n ⴝ 16)
Garlic
(n ⴝ 13)
Placebo
(n ⴝ 9)
p ANOVA
⫺12.4 ⫾ 44.6
⫺40.5 ⫾ 68.5
⫺2.3 ⫾ 30.8
1.1 ⫾ 30.2
⫺1.0 ⫾ 8.3
⫺5.4 ⫾ 7.0
10.4 ⫾ 4.2
3.3 ⫾ 31.2
3.1 ⫾ 11.0
0.2 ⫾ 3.6
0.5 ⫾ 4.0
⫺1.2 ⫾ 5.4
0.7 ⫾ 3.6
0.1 ⫾ 3.5
41.9 ⫾ 109.6
⫺25.2 ⫾ 83.6
1.3 ⫾ 17.0
⫺6.0 ⫾ 15.0
⫺0.1 ⫾ 9.3
⫺5.7 ⫾ 6.8
5.5 ⫾ 10.4
⫺5.1 ⫾ 19.8
⫺1.4 ⫾ 9.0
1.5 ⫾ 4.5
0.3 ⫾ 7.3
⫺0.4 ⫾ 7.9
⫺1.8 ⫾ 4.6
0.4 ⫾ 8.7
3.2 ⫾ 43.3
⫺58.0 ⫾ 79.8
⫺2.8 ⫾ 22.9
⫺4.3 ⫾ 21.3
0.8 ⫾ 4.7
1.7 ⫾ 7.1
⫺1.6 ⫾ 14.2
⫺4.4 ⫾ 19.0
⫺7.7 ⫾ 17.3
0.5 ⫾ 5.4
1.5 ⫾ 4.4
⫺2.5 ⫾ 6.8
⫺0.9 ⫾ 5.2
1.1 ⫾ 8.0
⫺14.8 ⫾ 57.2
⫺24.1 ⫾ 98.3
1.0 ⫾ 23.3
3.3 ⫾ 21.3
0.7 ⫾ 8.6
⫺0.1 ⫾ 6.0
2.3 ⫾ 11.8
⫺10.8 ⫾ 49.6
0.6 ⫾ 6.6
⫺0.4 ⫾ 7.2
⫺2.2 ⫾ 9.0
4.0 ⫾ 7.0
1.3 ⫾ 10.2
⫺2.6 ⫾ 6.8
0.17
0.70
0.96
0.61
0.94
0.01
0.14
0.75
0.16
0.81
0.60
0.19
0.56
0.68
Apo ⫽ apolipoprotein; HDL ⫽ high density lipoprotein; LDL ⫽ low density lipoprotein; Lp(a) ⫽ lipoprotein; TG pp ⫽ triglycerides postprandial.
LDL diameter in angstroms.
JACC Vol. 35, No. 2, 2000
February 2000:321–6
diastolic blood pressure. The only significant finding was a
greater reduction in LDL peak particle diameter in the
pattern A garlic and placebo groups compared with the
pattern B group. This difference in LDL peak particle
diameter was not paralleled by significant change in triglycerides or HDLC which is often reported with LDL peak
particle diameter change. In addition, there was no difference in the LDL percent distribution in the seven LDL
subclass regions. In particular, there was no significant
difference between placebo and treatment for postprandial
triglyceride concentration. Elevated fasting triglycerides
(⬎160 mg/dl) and low HDLC (⬍35 mg/dl) are often, but
not always, associated with LDL pattern B. In this investigation, 48% of LDL pattern B subjects had a fasting
triglyceride ⬍160 mg/dl and only 11% had a HDLC ⱕ35
mg/dl. Forty-four percent of LDL pattern B subjects would
have been mistakenly classified if fasting triglycerides ⬎160
mg/dl or HDLC ⱕ35 mg/dl were used as surrogate markers
of LDL subclass pattern B.
Previous garlic studies. It is important that no diet composition or BMI changes were documented in this investigation since changes in these parameters could have affected
lipoprotein values independently of any garlic effect. A
previous garlic study in 42 subjects reported an 8% LDLC
reduction in response to 900 mg/d for 12 weeks compared
with placebo, which was statistically significant (p ⬍ 0.05)
(17). A second investigation in 41 subjects reported a
significant reduction (7%) in total plasma cholesterol attributed to an aged garlic extract (4). However, in both studies,
subjects were either questioned as to any changes in diet or
advised to adhere to a National Cholesterol Education
Program Step I diet. There was no report of quantitative
diet record assessment. Thus, it is unclear if diet change
could have accounted for part of the observed LDLC
lowering result in the garlic group. The reports using
metaanalysis to assess the effect of garlic on lipoproteins do
not include diet records as a criteria for inclusion in their
analysis. A recent investigation of a garlic oil preparation
used seven-day diet records to assess nutritional status and
found no change in total cholesterol, LDLC, HDLC or
triglycerides in 25 subjects in a randomized, placebo controlled design in which it was documented that no diet
composition change occurred (5). The lack of LDLC
change in this recent study and the study reported here may
be due to inadequate sample size although this is unlikely
since the previous studies with 42 and 41 subjects reported
a significant reduction in total cholesterol and LDL cholesterol (4,17).
Apo E isoforms. Another potential confounding variable
in LDLC response to diet change is Apo E isoform
differences. Individuals with an E4 allele have a significantly
greater reduction in LDLC than individuals with the Apo E
3/3 isoform in response to a reduced fat diet and differences
in the prevalence of the E4 allele in the treatment and
placebo groups may confound data interpretation (30). In
Superko et al.
Garlic, Lipids, LDL and HDL Subclass Distribution
325
our investigation the E4 allele was almost equally distributed, five in the garlic group and seven in the placebo group.
It is unlikely that diet response differences due to Apo E
isoform differences contributed to any bias in our investigation.
Postprandial lipemia. In our investigation, both the placebo and treatment group showed reduced postprandial
triglycerides at the second test. Following garlic treatment,
the postload triglycerides decreased 49.6 ⫾ 73.6 mg/dl in
the garlic group. When the postprandial triglyceride reduction was compared within the garlic group for treatment
effect versus baseline, a significant difference (p ⬍ 0.003)
was found, but when compared with the 24.8 ⫾ 87.1 mg/dl
reduction in the placebo group, the postprandial triglyceride
reduction in the garlic group lost its statistical significance
(p ⫽ 0.28). Improvement in postprandial lipemia has been
reported in a study group of 24, but the change was not
statistically significant compared with the control group (2).
Conclusion. A recent randomized trial has reported no
effect of garlic oil on lipoprotein cholesterol levels (31). The
double blind, randomized clinical trial we report confirms
this lack of effect of garlic on routine measures of triglycerides, total, LDL and HDL cholesterol. However, it
contributes new information to the field by further revealing
no significant effect of 300 mg t.i.d. of KWAI garlic on
LDL peak particle diameter, LDL subclass distribution,
Apo B, Lp(a), HDL subclass distribution and postprandial
lipemia. Our study controlled for potential variables including diet composition, BMI and Apo E isoforms. There was
a significantly greater reduction in LDL peak particle
diameter in LDL pattern A subjects who received either
garlic or placebo compared with LDL pattern B subjects
treated, and the implications of this are unclear.
This issue is clinically relevant since lipid lowering treatments such as low fat diet, nicotinic acid, gemfibrozil and
hormone replacement therapy have been shown to have an
effect on either lipoprotein subclass distribution, Lp(a) or
postprandial lipemia that is not apparent from routine
measures of triglycerides, total, LDL or HDL cholesterol
(7–11,32). For example, change from a 46% fat diet to a
24% fat diet resulted in a 10% reduction in LDLC in LDL
pattern A men and 20% in LDL pattern B men (p ⬍ 0.02)
(8). No such differential response was documented with
garlic in this investigation. Differences in LDL subclass
pattern and response to treatments are important in predicting CAD risk and in selecting the most appropriate
treatment (7). Potential antiatherogenic mechanisms not
tested in this trial may contribute to an, as yet, unclear role
for garlic in the atherosclerosis armamentarium (33 –35).
Acknowledgments
We thank Patricia Blanche for her help in laboratory
organization and the technical assistance of Laura Holl,
Bahareh Sahami and Charlotte Brown. We would also like
326
Superko et al.
Garlic, Lipids, LDL and HDL Subclass Distribution
to thank Robin Rawlings, RN, for clinical support and Paul
T. Williams, PhD, for statistical advice.
Reprint requests and correspondence: Dr. H. Robert Superko,
Berkeley HeartLab, 1875 South Grant Street, Suite 700, San
Mateo, California 94402.
REFERENCES
1. Silagy C, Neil A. Garlic as a lipid lowering agent—a metaanalysis. J R
College Physicians 1994;28:2– 8.
2. Warshafsky S, Kamer RS, Sivak SL. Effect of garlic on total serum
cholesterol. A metaanalysis. Ann Intern Med 1993;119:599 – 605.
3. Rotzsch W, Richter V, Rassoul F, Walper A. Reduction in postprandial lipemia caused by Allium Sativum. Arzneimittel-Forschung/Drug
Research 1992;42:1223–7.
4. Steiner M, Khan AH, I-San Lin R. A double-blind crossover study in
moderately hypercholesterolemic men that compared the effect of aged
garlic extract and placebo administration on blood lipids. Am J Clin
Nutr 1996;64:866 –70.
5. Berthold HK, Sudhop T, von Bergmann K. Effect of a garlic oil
preparation on serum lipoproteins and cholesterol metabolism. JAMA
1998;279:1900 –2.
6. Krauss RM. Heterogeneity of plasma low-density lipoproteins and
atherosclerosis risk. Curr Opin Lipidol 1994;5:339 – 49.
7. Superko HR. The atherogenic lipoprotein profile. Science Med
1997;4:36 – 45.
8. Dreon DM, Fernstrom HA, Miller B, Krauss RM. Low density
lipoprotein subclass patterns and lipoprotein response to a reduced-fat
diet in men. FASEB J 1994;8:121– 6.
9. Superko HR, Krauss RM. Differential effects of nicotinic acid in
subjects with different LDL subclass patterns. Atherosclerosis 1992;
95:69 –76.
10. Superko HR, Blanche P, Holl L, Orr J, Shoenfeld MJ, Krauss RM.
Reduction of plasma LDL and apo B levels with combined estrogen ⫹
progestin therapy in postmenopausal women is greater in women with
dense rather than buoyant LDL. Circulation 1998;98:I–7.
11. Superko HR, Krauss RM, DiRicco C. Effect of HMGCoA reductase
inhibitor (fluvastatin) on LDL peak particle diameter. Am J Cardiol
1997;80:78 – 81.
12. Krauss RM, Lindgren FT, Williams PT, et al. Intermediate-density
lipoproteins and progression of coronary artery disease in hypercholesterolaemic men. Lancet 1987;2:62–5.
13. Miller BD, Krauss RM, Cashin-Hemphill L, Blankenhorn DH.
Baseline triglyceride levels predict angiographic benefit of colestipol
plus niacin therapy in the Cholesterol–Lowering Atherosclerosis Study
(CLAS). Circulation 1993;88:I–363.
14. Watts GF, Mandalia S, Brunt JN, Slavin BM, Coltart DJ, Lewis B.
Independent associations between plasma lipoprotein subfraction levels and the course of coronary artery disease in the St. Thomas’
Atherosclerosis Regression Study (STARS). Metabolism 1993;42:
1461–7.
15. Miller BD, Alderman EL, Haskell WL, Fair JM, Krauss RM.
Predominance of dense low-density lipoprotein particles predicts
JACC Vol. 35, No. 2, 2000
February 2000:321–6
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
angiographic benefit or therapy in the Stanford Coronary Risk
Intervention Project. Circulation 1996;94:2146 –53.
Jain AK, Gotzkowsky S, McMahon FG. Can garlic reduce levels of
serum lipids? Am J Med 1993;94:632–5.
Nutritionist III. N2 N-Squared computing analytic software. Silverton,
Oregon 97381.
Superko HR. Effects of acute and chronic alcohol consumption on
postprandial lipemia in healthy normotriglyceridemic men. Am J
Cardiol 1992;69:701– 4.
Warnick GR, Albers JJ. A comprehensive evaluation of the heparinmanganese precipitation procedure for estimating high density lipoprotein cholesterol. J Lipid Res 1978;19:65–76.
Myers GL, Cooper GR, Winn CL, Smith SJ. The Center for Disease
Control-National Heart, Lung and Blood Institute Lipid Standardization Program: an approach to accurate and precise lipid measurements. Clin Lab Med 1989;9:105–35.
Young SG, Smith RS, Hogle DM, Curtiss LK, Witztum JL. Two
new monoclonal antibody-based enzyme-linked assays of apolipoprotein B. Clin Chem 1986;32:1484 –90.
Corsetti JP, Sterry JA, Sparks CT, Weintraub M. Effect of weight loss
on serum lipoprotein (a) concentrations in an obese population. Clin
Chem 1991;37:1191–5.
Weisgraber KH, Rall SC, Mahley RW. Human E apoprotein heterogeneity. J Biol Chem 1981;256:9077– 81.
Zannis WI, Breslow JL. Human very low density lipoprotein apolipoprotein E isoprotein polymorphism is explained by genetic variation
and posttranslational modification. Biochemistry 1981;20:1033– 41.
Cooper GR, Smith SJ, Wiebe DA, Kuchmak M, Hannon WH.
International survey of apolipoproteins A-I and B measurements
(1983–1984). Clin Chem 1985;31:223– 8.
Krauss RM, Burke DJ. Identification of multiple subclasses of plasma
low density lipoproteins in normal humans. J Lipid Res 1982;23:97–
104.
Nichols AV, Krauss RM, Musliner TA. Nondenaturing polyacrylamide gradient gel electrophoresis. In: Segrest JP, JJ Albers, editors.
Methods in Enzymology: Plasma Lipoproteins. New York: Academic
Press, 1986:417–31.
Krauss RM, Blanche PJ. Detection and quantitation of LDL subfractions. Curr Opin Lipidol 1992;3:377– 83.
Feldman DS, Gagnon J, Hofmann R. StatView II. Simpson Brain
Power Inc. 1994.
Lopez-Miranda J, Ordovas JM, Mata P, et al. Effect of apolipoprotein
E phenotype on diet-induced lowering of plasma low density lipoprotein cholesterol. J Lipid Res 1994;35:1965–75.
Berthold HK, Sudhop T, Bergmann K. Effect of a garlic oil preparation on serum lipoproteins and cholesterol metabolism. JAMA 1998;
279:1900 –2.
Kim CJ, Jang HC, Cho DH, Min YK. Effects of hormone replacement therapy on lipoprotein (a) and lipids in postmenopausal women.
Arterioscler Thromb 1994;14:275– 81.
Phelps S, Harris WS. Garlic supplementation and lipoprotein oxidation susceptibility. Lipids 1993;28:475–7.
Das I, Khan NS, Sooranna SR. Potent activation of nitric oxide
synthase by garlic: a basis for its therapeutic applications. Curr Med
Res Opin 1995;13:257– 63.
Orekhov AN, Tertov VV, Sobenin IA, Pivovarova EM. Direct
antiatherosclerosis related effects of garlic. Ann Med 1995;27:63–5.